Syed Muzaffar J

Developed robust forecasting engines using TensorFlow RNN, Prophet, and AWS forecasts for Area Sales Managers. Built an ensembled forecasting engine to predict sales across multiple African countries. Managed projects including recommender systems, promo simulators, and decision boards. Developed end-to-end pricing solutions using neural networks (Keras) for daily pricing recommendations across 1.5 million store-SKU combinations. Assisted clients in measuring promotion effectiveness and developed order monitoring systems to track purchase patterns and flag suspicious orders.

Diagnosed data pipelines and rationalized reporting, reducing total reports consumed from 250 to 13. Led the development of optimal pricing strategies using price elasticity and cannibalization models for a major UK retailer. Developed capacity forecasting models using LARS, Random Forest, and XGBoost, alongside Monte Carlo simulations for inventory optimization. Developed customer segmentation algorithms to identify various customer buckets for a global payments and technology company.